CHAPTER 11
PHYLLIS CULHAM
IMPERIAL legions ordinarily maintained a standard marching order in theater, although that could easily be modified by the theater commander for local circumstances. Josephus, for instance, describes Vespasian’s advance with three legions and a total of sixty-six thousand combatants from the coast into Galilee in 67. Lightly armed and armored auxiliary forces preceded so that they could rapidly deploy on the tactical offense to meet any attempt at harassment. They pursued attackers even into woods, which, as noted below, challenged the abilities of legions to maintain their order. These light and missile-wielding forces were followed by mounted and unmounted Romans, presumably to provide cover for those next in the column, if the skirmishers had gone off in pursuit. They were followed by units which might be needed to clear a route for the advance or to build a camp in the face of resistance, including surveyors to straighten routes and level campsites. Then cavalry escorted the equipment of the generals; maps, documents, furniture for the praetorium in a camp, even the payroll were presumably included there. Vespasian himself came next in the column, escorted by specially selected foot, horse, and spearmen. Mules pulled rams and artillery. Then came the legionary commanders and prefects with more picked troops accompanying the legionary standards, followed by the trumpeters. Next the legions themselves were marching ahead of their servants and baggage with mules. They were followed in this particular case by mercenaries, behind whom trailed a rear guard of more Roman infantry and cavalry, presumably detailed from legions (Joseph. BJ 3.115–126).
Auxiliaries usually supplemented legions as lighter and more mobile forces. Put less kindly, they were likely to bear the first contact with the enemy and to screen legions. Conservation of force was a hallmark of Roman Imperial warfare, and the necessity for engaging legions in combat was always seriously weighed. Mere mercenaries were not left in the back of that column for their protection but because they were not relied on in an emergency. Rams and artillery were included routinely in such an order of march, since, as explained below, the usual grand tactical (theater-level) assumption was that the war itself was likely to resolve into a series of sieges.
Figure 11.1 Tombstone of Marcus Caelius, a centurion of the Eighteenth Legion killed in the Teutoberg Forest battle (near Kalkriese, Germany) along with his two servants. His monument is one of the few written records attesting the great Roman defeat. Landesmuseum Bonn. Photo Credit: L. Tritle.
The road builders and surveyors were toward the front of the column for a reason. A great premium was placed on the ability to move along with the artillery, some of which might even have been needed on camp walls in hostile territory, and on the ability of the legions to march in good order, ready to form fighting units. Vespasian’s advance required the leveling of a mountainous route from Gabara into Jotapa at the rate of six miles in four days (Joseph. BJ 3.35). Varus’s disastrous march through what he believed to be allied territory in Germany required leveling and building. This required cutting trees, building bridges, and clearing roads even before his strung-out forces reached the dense Teutoburg forest in which the column was ambushed (figure 11.1). He was not advancing in regular order, and combatants were mixed with noncombatants and vehicles in such way that they could not form into units, exactly what Arrian tried to prevent later (Dio 56.20.2–21.1; Joseph. BJ 3.91, 93 on general necessity to keep the combatants clear of the rest when advancing). Varus was able to construct a camp while under attack, although it had to be adapted to the terrain. Varus supplies a worst-case scenario for an advance, just as Vespasian’s advance into known, hostile territory can serve as the state-of-practice model (Onos. 6–9 emphasizes the commander’s personal responsibility for order in a march).
The formula could vary, as Germanicus illustrated east of the Rhine a few years later in 14, when that territory was better understood. Of course, he bridged the Rhine before crossing with about fourteen thousand men; then he built a camp on a site already used by Tiberius. He set out from that down a long, rough route less likely to be watched by Germans who knew how Roman columns needed to advance, even without artillery, sending a subordinate ahead with auxiliaries to clear a marching route through the woods. He did not hesitate to divide his forces to wreak terror on a broader front. His marching order on the way back was informed by his knowledge that ambushes awaited in the woods. Cavalry and auxiliaries again led the march, with three legions advancing abreast and one behind, in something resembling a battle order rather than a column. The legions were followed by more auxiliaries in a sandwich arrangement preserving the legions as an asset. They met no resistance until, as in the case of Varus, the forested land had caused gaps in their order. The main attack was directed at auxiliaries in the rear, but the legion on the left of the advance was able to turn away the attack, while the forces in the lead were able to emerge from the woods and establish a camp (Tac. Ann. 1.48–52). This incident highlights both reliance on construction capabilities even under the wildest conditions as well as adaptability in maintaining unit cohesion and tactical deployment.
In another instance of adaptability, when Germanicus set out to attack the Chatti in the next year, leaving with light arms and auxiliaries so that he could move rapidly, he had a subordinate build roads and bridges after him, since he anticipated returning in a rainy season (Tac. Ann. 1.56.1). In both of these seasons, Germanicus’s targets in his advance had been German villages and their populations. Terror was often an effective force multiplier, since it did not risk legionaries in battle and might obviate resistance even before it appeared. Germanicus probably also intended to shore up his men’s morale (he had some of Varus’s survivors in his army) by allowing sheer revenge, which might also have aided him at the strategic level by deterring further actions against Roman interests.
Arrian’s advance to confront invaders in the mid-130s, in Cappadocia in Asia Minor, differed slightly from Germanicus’s two previous advances into hostile territory, yet a similar pattern still appears. The differences between Arrian’s advance and Vespasian’s, besides those of scale, are mainly that he had even more auxiliaries out in front and counted on German horsemen to support skirmishers rather than on Roman cavalry, and Arrian did not bring engineering units in his effort to catch mounted invaders. The core of the advance still sounds familiar: selected Roman units, Arrian with elite cavalry, artillery, standards, legion. At the tail a provincial militia replaced mercenaries, and the baggage train was sandwiched between auxiliary units (cf. Bosworth 1977: 251). Arrian claimed to see it as his primary function (and that of his centurions) during the advance to encourage order and prevent any disorder. Arrian’s most detailed planning addressed the problem of how to preserve order and avoid risk to Roman forces pursuing a routed enemy. He also allowed for the problem which typically confronted Romans: attempted encirclement by tactically more numerous forces; he was going to respond with cavalry and archers. His planning for the expected rout, however, reached as far as considering how the cavalry might change to fresh horse during the pursuit and destruction of the invaders (Arr. Scripta Minora 1895: 80–5; translated: The Roman Army Page http://members.tripod.com). In short, legions in the Principate were highly protected assets, even during an aggressive advance. Romans preferred to advance in a standardized column yet were capable of highly adaptive tactical deployment.
CAMPAIGNS AND ASYMMETRICAL WARFARE
Varus and Germanicus also point to the topic of the obvious danger during an advance: attempts at ambush, not merely harassment. Roman commanders in theater must have viewed this with ambivalence. Armies of the Principate more often had to contend with the problem of catching an enemy to defeat. They rarely encountered opponents willing to face them symmetrically in unit-on-unit combat in battle lines. Roman legions were able to win asymmetrical encounters because they were supplemented by more flexible, sometimes local, auxiliary units and sometimes Roman allies; because they kept legions or cohorts from legions in reserve for tactical deployment at critical junctures; and because they were able to maintain unit cohesion which made it very difficult to defeat them tactically in any significant sense. Varus’s great defeat illustrates the penalty of failing to maintain unit cohesion.
Roman commanders usually sought combatant-on-combatant battle, since that was to the Roman advantage, unless they were extremely outnumbered. For instance, in 16 Germanicus was pursuing Arminius, who had designed the massacre of Varus’s legions. Germanicus’s campaign illustrates both how opposition who did not intend to engage in battle might do so anyway and why Roman legions had to be ready for a variety of combat conditions in one campaign. Germanicus attempted to eliminate much of the requirement for construction before his advance in that year by taking ships up the Ems River. Nevertheless he still landed short of his target and had to build bridges for the infantry, although the cavalry could often use fords.1 Arminius was a special, hard case who knew well Roman strengths and preferences. In the previous season, he had used woods for ambushes and attempted to force the Romans into bogs, but he was capable of more openly aggressive behavior. He originally intended a night attack on Germanicus’s camp, but that was anticipated, and he made no attempt on a camp on alert. With other chieftains and their forces gathered, he had no choice but to stand and fight to justify his standing in his own coalition.
It is a measure of how much battle per se favored Romans that Germanicus even let Arminius choose the ground, Idavisto, a plain on the Weser River, with hills and forests on the other sides. Arminius preset tactical dispositions with men on the plain and the edges of the forest, placing himself on the hill, ready to come rolling down when it would do the most damage. Germanicus predictably sent auxiliaries into first contact in a marching order ready to deploy as a fighting front. Bowmen followed, although later Roman preference was for archers behind Roman legions to provide covering fire. When Arminius’s own force charged down very early, Germanicus had time to order the chosen cavalry, which typically accompanied the commander, to attack the flank Arminius presented to the Roman advance and to order the rest of the cavalry to ride up the hill behind him. Arminius escaped wounded; German forces were slaughtered.
Nonetheless, tactical victory on one plain did not give Romans control of the area. Enraged, grieving Germans, including some who had not been previously hostile, inflicted many casualties on the Romans’ departing marching order. Hastily assembled Germans tried a trap at a narrow, swampy space between a river and a forest with an earthwork on one side, meant to mark a tribal frontier. But masses of Germans in a small clearing were unable to take advantage of numbers by charging with force; nor could they effectively wield weapons which were much bigger than the short Roman sword; nor could they defend themselves against legions in close combat without helmets and armor (Tac. Ann. 2.5–23). The first engagement argued against fighting where Roman cavalry wings could use their mobility; the second argued against fighting legions densely confined. One of the biggest differences between the armies of the Principate and those of the Punic Wars and mid-Republic was the solid performance of Roman cavalry (such as the Batavians, on which see below) attached to each legion.
Nothing about these encounters encouraged subsequent enemies to oppose Romans openly in battle. Nor could Romans believe that even a complete tactical destruction of enemy combatants in the field would give them control of an area or even the ability to pass through it safely. Through Germanicus’s efforts, Romans secured the symbolic victory of digging up one of Varus’s lost standards; prisoners supposedly reported that Germans feared Romans and thought they were unbeatable (Tac. Ann. 2.25.4). Revenge, honor, and terror had probably been the Roman strategic objectives all along, and battle aided mainly the first and especially second of those. Arminius was the first example of a problem that emerged infrequently but always dangerously in the Principate, namely, what today would be called blowback. Varus considered Arminius a personal friend and would not listen to warnings about him, as he had long advised the Romans about other Germans (Dio 66.19.2–3). If such a Roman “friend” or client turned back around, he brought a more sophisticated understanding of Roman warfare to subsequent combat, for example, Arminius’s respect for Roman marching camps and ability to build something like them himself; note too his keeping of reserves. Romans courted and rewarded local elites and preferred client princes on their frontiers because they were low-cost force multipliers. Clients also supplied additional capacities in asymmetrical conflict. There were dangers to that practice, but Germanicus’s campaigns to exterminate potential opposition did not Romanize the region thoroughly either. Germanic levies were trusted sufficiently to catch Germans raiding across the Rhine later in 50, although the Roman commander was unable to trap the marauders into a battle with legions (Tac.Ann. 12.27.3–4). Britons proved more prone to tackle legions in battle.
Certainly, Claudius’s incursion into Britain did not accomplish control of the population or even unchallenged ability to live in and move about the territory. As early as 47, the allegedly allied Iceni were leading other tribes in resistance, albeit not in battle, thinking Romans would not fight in winter. They took a (familiarly sounding) stand behind earthwork with narrowed approaches in one case and otherwise refused open engagement, attempting ambush instead. The Roman governor, P. Ostorius Scapula, went on to set permanent camps among other tribes (Tac. Ann. 12.31–32). These represented grand tactical, aggressive positioning in an attempt to train allies and erode a resentful population’s will to resist. The result among the Silures was the emergence of Caractacus as a coalition leader. He took a stand in battle behind allegedly improved stone and earthworks which could not withstand a Roman unit hacking on it in testudo formation (Tac. Ann. 12.33, 35). After Caractacus, the Silures fell back on even less symmetrical tactics of picking off encamping legions and foragers. Even Roman “picked men” could be caught in those situations. Although Ostorius consistently tried to conserve force and rely on first use of auxiliaries, they were not always up to the job (Tac. Ann. 12.39).
Perhaps the fact that Romans overcame the traditional hill forts so easily, along with confidence in fervor and numbers, influenced Boudica of the Iceni in 61. She had been able to gather forces in the absence of the governor, Suetonius Paulinus. Her initial targets were Roman veteran settlements like Camulodunum, which were indeed instruments of Romanization and Roman control. Camulodunum relied on local false friends, little Arminiuses, and the Ninth Legion was almost wiped out while attempting relief (Tac. Ann. 14.32). Paulinus, we are told, was atypically reluctant to engage in battle, since he “feared their numbers and their insanity.”2 Sources tell that the Britons were torturing and taking no prisoners, also emphasizing that their attacks focused on economic sites rather than outnumbered Roman units (Dio 62.8.1; Tac. Ann. 14.33.4–6). Probably Paulinus met Boudica on a road toward her next target. He was unable to extend his line far enough to counter (allegedly) 230,000 Britons. Tacitus attributes to Paulinus a simple, memorable set of instructions: hurl javelins, knock down opponents with shield boss, kill with sword; do not try to collect loot. While Boudica’s forces sang triumphantly, Roman forces advanced in good order and then charged. Of course, they broke into the less practiced mob and, equally predictably, were surrounded by forces with various weaponry, including chariots. Roman order eventually prevailed (Tac. Ann. 14.34).
Nonetheless, Dio claims that revolt was ended not by success in battle, since survivors were re-forming, but by the death of Boudica soon afterward (Dio 62.12; cf. Tac. Agr. 16). Although Paulinus unleashed a campaign of terror not only upon hostile tribes but on the unaffiliated, and famine was severe because the war had precluded harvest, Nero still blamed him for delay in the war. Tacitus claims that his successor simply ended hostilities and called it peace (Tac. Ann. 14.39). Nor was it easy to maintain Roman morale on the British frontier. Dio’s account of Claudius’s initial advance into England stresses that Romans resented the driving rain and that Roman soldiers feared a trip to England as a departure from the oikoumene, the known world. Predictably, Plautius had trouble simply finding the opposition and then had to contend with rivers, lakes, and swamps. So much water left Mediterranean forces disgruntled (Dio 60.9.3; 60.19.1–5; cf. Tac. Ann. 1.30.1–3 in which Roman forces take a hard, winter rain as divine punishment for a mutinous mood).
Tacitus makes that very distinction between overrunning an area and actually ending opposition there when discussing his father-in-law, after whom he named the Agricola (Agr. 17). In the Principate, the smallest unit one could evaluate to see if an objective had actually been obtained was the campaign and not the relatively infrequent battle. In Agr. 18, Agricola set out in 75 with a classic carrot-and-stick campaign of unprovoked destruction followed by the practice of clementia, namely, acceptance of hostages and pledges from peoples who did not want to be next. Seasonal campaigning against Roman winter camps was subdued via the persistence of terror inflicted in the summers. When Agricola came across an exiled Irish prince, he “befriended” him with an eye to subsequent campaigns on that island. He made effective use of a fleet for power projection along untouched coastlines. Agricola met the threat of a joint assault by tribes who outnumbered him by moving forward to meet them, and they responded by reducing their plans to a night attack on a Roman camp (Agr. 22, 24–6).
The next campaigning season anomalously led to battle. Agricola set out again with both fleet and army on a campaign of terror, only to find that he had provoked so much resistance that more than thirty thousand men were waiting across his line of advance on Mons Graupius. Agricola set a camp and placed the legions in front of it to preserve a chance of retreat. Trusted British auxiliaries faced the danger as the center in front of the legions. The opponents were on a hillside on the other side of the small plain, while their light charioteers dashed about in front of them in maneuvers meant to intimidate. Agricola did what most commanders of the Principate were to do in similar situations; he extended his line, so thinly that some staff officers wanted him to bring the legions into the line and not keep a reserve (Agr. 35). The actual battle began with an exchange of missiles, and Agricola ordered the outstanding Batavian auxiliaries forward. They were wholly successful in pushing opposition back up the hill, and Roman cavalry completely defeated the light chariots but could not sweep on up the hill on the rough ground. Massed forces began to descend from the hill and enveloped successful Roman auxiliaries driving forward, but Agricola’s reserved cavalry enveloped the mass now fighting on the plain. Survivors who reached the woods were able to turn and ambush the first pursuit. Agricola withdrew to winter quarters in a slow, systematic marching order, simply to show that he had no fear (Tac. Agr. 35–9).
This illustrates again how little Roman opponents profited from the relatively infrequent battle, even when they picked the ground, often choosing hillsides for gathering momentum. The engagement also demonstrates their usual preference for attacking Roman armies when they were on the defensive (e.g., encamped at night), hoping for confusion which might disrupt unit cohesion (a trend which began with Gauls against Caesar). Even in these British campaigns, the Romans’ Mediterranean habit of seasonal campaigning seemed intractable. It encouraged resistance by Britons who believed, rightly or wrongly, that Roman forces were more likely to respond with less force in winter. Persistence of this habit might have been encouraged by another Roman predilection, namely, trying to ensure that Roman forces were reasonably comfortable and fresh. Auxiliaries played their usual role of bearing the first shock. A Roman author ascribes the domination and deterrence of the enemy through fear not to tactical victory in battle but to the cumulative effect of campaigns of terror. It is not surprising that commanders who rely militarily on terror consider morale of their own forces important.3 Both Romans and their northern opponents tutored one another with each contact. Romans after Varus were hyper-organized while advancing and well-prepared to move from marching to fighting order. Northern opponents responded to long-term hostilities, including campaigns of terror against local populations, by coalescing behind dynamic leaders who often led them directly against Roman armies. Romans in turn developed a repertoire of responses to numerically superior enemies and to local earthworks yet remained highly adaptable to local conditions.
CAMPAIGNS AND SYMMETRICAL WARFARE
The ultimate potential of Roman legions against an equally trained, symmetrical enemy could emerge most clearly in civil war. Appian had already claimed that, at the battle of Pharsalus in 48 BC, Greek units fled the field in horror at the ferocity of legion-on-legion combat, both sides killing with the short sword (B Civ 2.11.79). In the civil war of 68/9, the battles at Cremona, near a confluence of strategically significant, northern Italian roads, must have been equally terrifying. The first battle, between forces loyal to the contender Vitellius and those of the short-lived Otho, illustrated two different approaches, the innovative and the conservative. Caecina for the Vitellians had planned to lure pursuers into a trap, but his overconfident forces approached too close, when no one rushed out against them, and were surrounded themselves. Paulinus on Otho’s side showed extraordinary devotion to the by-the-book Roman advance; he stopped pursuit to fill in ditches and to level the ground! This also demonstrates how fast his engineering units must have been able to work in the vineyards and fields around Cremona.
As Caecina’s men fled, lines broken, Paulinus sounded retreat, fearing that more Vitellian forces would arrive fresh from Cremona and that his own tired men could not withstand much more if caught in pursuit. Paulinus represented perhaps one extreme on the relatively short Roman spectrum of circumspect, controlled deployment, keeping forces fresh versus daring deployment and maneuver. Tacitus reports that some armchair commentators approved, but Paulinus’s men did not (Hist. 2.24–26; 25 allows him the epithetcunctator, the “delayer,” in describing his devotion to planning and reason). The ability to order an end to pursuit of a rout was as remarkable in antiquity as the fact that Paulinus chose to do it. The contemporary Josephus was certainly struck by Roman unit cohesion so strong that they retained a capacity for maneuver even in the midst of battle (BJ 3.88, 105).
The second battle of Cremona was between Vitellian forces and legions loyal to Vespasian after the death of Otho. Antonius Primus, for Vespasian, encountered the Vitellian legions earlier than he anticipated and had to recall foragers with a signal, and some cavalry arrived frightened and disordered from a skirmish. Disorder spread, especially since units were jostling each other on the narrow road (perhaps justifying Paulinus in the first battle). Antonius threw his spear through a fleeing standard bearer, catching up the standard himself and turning it to face the enemy. About a hundred witnesses were stricken with shame and turned with him.4 His men realized that the narrow roads prevented escape, formed themselves up, and drove away the overconfident, increasingly disordered Vitellians (Tac. Hist. 3.16–18). The second battle of Cremona demonstrates the effects of both cohesion and lack of cohesion. Even when a clear, legal command structure was absent, veteran legionaries, realized (with reminders) that safety lay in unit cohesion as well as their ability to generate such cohesion en masse from their small units. It even demonstrates their ability to respond to a signal in dangerous and deeply distressing circumstances. Since this was a Roman-on-Roman action, the description cannot simply be counted as a literary topos of unique Roman discipline over the fabled indiscipline of barbarians.
A contemporary, dangerous conflict on the Rhine exhibited both symmetrical and asymmetrical features. This devolved from a nightmarish blowback situation in which individual, well-trained Roman auxiliaries, loyalty to their own peoples, became players in the Roman civil war.5 Other auxiliaries defected to Civilis throughout the conflict, often in the midst of battle with dangerous consequences to Roman units, although some Romans were able to maintain their unit’s integrity even when Batavians turned on them suddenly. Civilis’s initial successes fed even higher expectations, and Roman morale sank in correspondence to how well Civilis was doing. Civilis was as sophisticated as Romans in attracting allies with a combination of force against resistors, bribery, and selective clemency. He paraded captured Roman standards for audiences of both non-Romans and Romans, the latter discombobulated by seeing ranks of modified Roman standards in the opposition alongside what Tacitus describes as tribal animal totems (demonstrating that Roman auxiliary units had such group identity that they would keep their Roman unit’s standard even in revolt, only removing the imperial portrait). Batavians used complex tactical formations against well-defended Roman camps and ran legions off the field at Bonn and Vetera; they later built siege engines to attack Vetera and Gelduba, although they were outdone by more sophisticated Roman engineers (Tac. Hist. 4.30). Tacitus even describes Civilis as leader of a “real army” (Hist. 4.21). Predictably he attracted trans-Rhine German support, although his forces exhibited particular hostility toward Romanized peoples of German descent.
This blowback-induced conflict drew Gallic combatants into its draft. Julius Classicus was, predictably, a descendant of royalty among the Treviri. He joined a relief expedition to Vetera only to defect when he got there. Civilis himself began to worry about the rise of a Gallic confederacy (Tac. Hist. 4.61). Classicus displayed captured Roman standards along with Gallic tribal pennants; he also donned the insignia of a Roman commander (Tac. Hist. 4.59). Q. Petilius Cerealis probably had multiple reasons for sending Gallic levies home, saying that legions could see to the defense of Gaul (Hist. 4.71). Not only might the Gauls have been liable to turn in battle, this dismissal might also have been a morale-boosting display of confidence in the arrival of reinforcements, a show of generosity, and a fear of removing all men of military age from loyal communities. The war was finally resolved in a battle which began without Civilis’s personal presence. The Romans only defeated their own former auxiliaries who had breached their camp, when the Twenty-first Legion was able to find space to form up, to resist, and then to push out, another testimony to the significance of unit cohesion.
LONG WARS
Talented generals and perhaps a majority of imperial legions saw campaigns which did not feature battles nor even any clear outcomes or rewards. It is not surprising that morale became even more of an issue in such campaigns. Tacitus’s Annals offers a precise catalog of conditions Roman armies wished to avoid: absence of roads, lack of ports, ferocious chieftains, wandering peoples, country one could not live off. No less a concern were forces deteriorating through boredom, endangered if acting precipitously, given no credit for winning against “barbarians” but disgrace in defeat (12.20). Germany was such a frontier. Entanglement with German tribes even as allies sometimes opened new gates to conflict. Germans had always settled and resettled back and forth across the Rhine and were not inclined to defer to the Roman sense of a frontier. When Frisians in 58 resettled right on the Rhine where Romans wanted a demilitarized zone they began to plow without attacking Romans. Nero handed their chieftains citizenship and an eviction notice, and auxiliary cavalry was good enough to serve the eviction papers. But more Germans moved right back in, and their chieftain was an old ally against Arminius. A sadly sympathetic Roman governor had to dismantle their whole allied coalition in order to run them off (Tac. Ann. 13.54–56). The long-term consequences of Roman involvement in cross- and supra-Rhine shifting, German tribal alliances were becoming clear.
Tiberius spent much of his life and energy on two such frontiers, the Rhine and the Danube. According to Dio, problems on the German front pushed the Danube/Adriatic region into the next level of conflict in 6. Forces under Bato of Dalmatia looted the Greek cities of the Adriatic coast and marched on Roman Salonae once. They did stand and fight one legate, but otherwise it is not clear what their purpose or strategy was, despite that move toward Salonae and Bato the Pannonian’s attack on Sirmium. They kept themselves very lightly armed to move quickly and would not defend any territory or populations. They simply hid in their fortresses in the karst and raided from them. Resistance activities on this Adriatic front foreshadow Vespasian’s later difficulties in Judaea in distinguishing nationalist or other ideological resistance from brigandage. Augustus, as distant emperor, could not tell what was taking Tiberius so long and supposedly began to suspect that he might just want to retain a command under arms, which must have been immensely frustrating for Tiberius (Dio 55.29.–55.31.1).
Velleius Paterculus was a sympathetic comrade in Tiberius’s problems. He notes that the Pannonian Bato spoke Latin, understood Roman military discipline, and drilled his forces, illustrating how the Romans were always training potential opposition simply by exhibiting their disciplina. He also knew, as did Boudica later, to try to reduce colonies of Roman veterans early on, when raiding. This Bato also went in for ambushes, trying one on a full five-legion force. When Tiberius had to manage this theater later as emperor rather than as combatant, he did not think more forces covering more ground were the answer; he considered ten legions too big (Vell. Pat. 2.110–113). The “resolution” came under Tiberius’s successor when the population was demoralized by famine and disease, and the two Batos turned on each other.
Dio comments sourly from his vantage point more than a century and a half later that the supposedly pacified region was left with a brigandage problem, which “always” happened in such cases (Dio 55.34.7). He might have been thinking about such instances as Tacfarinas in Mauretania in 17 who appeared to be interested in both political resistance and profiteering from the beginning (Tac. Ann. 2.52.1–2). Predictably, he had served with Roman auxiliary forces; he even developed his own more lightly armed auxiliary forces. Marcus Furius Camillus was anxious to get him into battle from the beginning, which was not difficult given his confidence. After a defeat by Camillus, he became a more typical brigand for a while, although surrounding Roman cohorts and killing their commander in 20 was a mistake if he wished to remain a low priority criminal rather than an insurgent.6 He finally made the classic mistake of weighing himself down with too much loot. The base to which he retired was overrun, and he was driven into the desert.
He escaped attention for two more years until he demanded lands for settlement in a prefiguring of Marcus Aurelius’s problems on the Danube. A clever offer of amnesty to his men isolated him with a hard-core remnant. Romans used the technique of pre-positioning far forward camps from which to harass him; other Roman forces were deployed to defend economic targets like cities. Quintus Junius Blaesus actually decided to campaign through the winter. Tacfarinas fled into the desert after his brother was captured. To Tacitus’s disgust, Tiberius was again willing to consider his flight out of the imperium a solution and to reduce the number of Roman forces in the area (Ann. 3.74.5–7). Tacitus even used the phrase Long War (longum bellum) two years later when Tacfarinas reemerged and was able to attract other auxiliary units to join him, as Civilis could later. At that point, Tacitus ascribes to Tacfarinas the political rhetoric of revolt, speaking of liberty versus slavery (Ann. 4.24.1). King Ptolemy of Mauretania, a Roman puppet, supplied high-quality local forces for counterinsurgency and was greatly praised for it later. Finally, Tacfarinas was caught literally napping in the ruins of an old Roman camp, and Tacitus ascribes the consequent butchery to Roman resentment of hardship in that theater (Ann. 3.25.3). Yet a form of Long War persisted over a generation later in 69, when Tacitus comments in passing on the excellent military skills of some Mauretanian auxiliaries who had had much practice as brigands (Hist. 2.58). This case demonstrates why asymmetrical conflicts could not always be easily categorized as economic/criminal or political/military, and how it might have been reasonable to hope to turn some possible insurgents back to loyalty.
Marcus Aurelius’s activities in the 170s illustrate the nature of Long Wars most forcefully. He juggled the complex factors of military assets, raiders, migrants, and money on the Danube front with sophistication, aiming at the great goal of conservation of force. It was a constant guessing game as to what each people really wanted, and many did not know themselves. He agreed to some peoples’ desire to resettle. Some might have preferred to resettle but agreed to accept funds to be “friends” beyond the frontier. Of course, most wanted both lands and subsidies, but Marcus Aurelius had to exercise the art of the possible. Some peoples who were addicted to raiding might still be offered incentives, if he believed that they could be turned around against more dangerous peoples. He sometimes placed a cash bounty on especially nefarious leaders (Dio 72.1–14). Some like the Iazyges he would have preferred to “destroy,” but he agreed to a settlement which moved them back even further from the Danube (Dio 72.16.1). He followed precedents noted above in placing permanent camps right among the Marcomanni and Quadi in an attempt at controlling the region, and their attempt to move farther away demonstrates that the Roman encampments must have inhibited them.
In other words, Roman military deployment on the northern frontiers had to account for immigration control, refugee management, river patrol, and prevention of crimes against property. A case can be made that the limes system of roads, stockades, and towers was meant to prevent brigandage and to protect movement of persons and goods as much as for strategic defense against threats beyond the Rhine, Danube, and Euphrates (Shaw 1984: 12). Here, as in Britain and Africa, Roman forces might be relegated to dispersed, permanent camps among the most suspect peoples. Yet it was a mistake to assume that Roman legionaries were not always ready to fight and were not adaptable to environments much unlike the Mediterranean world, as the Iazyges found out when they turned for an offensive charge on the frozen Danube, thinking that Romans would not fight on ice. The Romans supposedly put down their shields to stand on and formed a compact unit, grabbing bridles and pulling down horses whose momentum would not allow them to stop. On foot, Roman armor won (Dio 72.7, plausibly vouched for by Tac. Hist. 1.79).
COLD WAR AND LONG WAR
The Parthian frontier presented a different kind of problem in asymmetric and Long Warfare. Roman emperors and Parthian kings generally observed political settlements respecting something like a Euphrates Line, but ambitious leaders on either side could upset understandings at will. Both sides found it generally advantageous to settle into Cold War punctuated by tests, raids, and attempts to co-opt buffer states, continuing a Great Game through cats-paws, proxies, and clients. The Parthian king could reach full mobilization only by calling upon other Parthian nobles for forces. Since the dry expanses of the Near and Middle East competed with the requirements of the Rhine and Danube fronts, force conservation, especially of Roman legionaries, was vital. The situation was sometimes simpler than that on the Danube, since Parthians often aimed expeditions at targets which could be anticipated by Romans. They would then calculate in ways understood by Romans, withdrawing when massed against so that risk outweighed possible gains. Tacitus claims that the great Cnaeus Domitius Corbulo once urged his forces on against this “wandering enemy, ready for neither peace nor war” (Tac. Ann. 13.39.14–15).
Corbulo arrived on the Parthian front in AD 58 with a reputation for disciplinary severity. Armenia, a perpetual soccer ball bounced between both sides, was claimed by the brother of the Parthian king, Tiridates. Typically, Corbulo very much wanted a battle; he could not force one. The term “Parthian shot” referred to their highly skilled tactic of firing back over the shoulder while avoiding close combat with armored Romans. Then they could regroup at a distance. Tacitus describes Corbulo as forced to follow the “example of the enemy” and to widen the scope of the war by dividing his forces (Tac. Ann. 13.37.2).
Corbulo was reduced to attacking stationary targets, namely, Armenian fortresses Tiridates could hole up in. He targeted Artaxata, the major city of Armenia, knowing it would make a problem for Tiridates: he could not afford to appear to leave a Roman siege undisturbed, nor did he want to engage with Romans in close combat. Tiridates used the standard Parthian tactics of seeming to advance against Roman forces, then seeming to withdraw. Corbulo had also read the playbook and was too careful to pursue. The next day Artaxata opened its gates and surrendered. Corbulo destroyed the site without doing violence to the residents, under his policy of destroying strongholds he could not hold (Tac. Ann. 13.41.3).
Tacitus claims that the destruction produced terror among Armenians, sending much of the population fleeing into caves or hills, so Corbulo set out to evoke more fear by taking Tigranocerta. Corbulo’s forces spent the heat of summer showing sympathy to those who surrendered, chasing those who ran, burning fugitives out of caves, and harvesting local crops rather than in battle. Tigranocerta finally surrendered, saving its citizens’ lives and property.7
When Corbulo had to see to the survival of the puppet regime in 62, he articulated the unusual nature of the Cold War by saying that he would rather “have a war than make war,” instructing his commanders not to rush to the front (Tac. Ann. 15.3.1). Instead he seized his pen and wrote the Parthian king Vologeses telling him to break the siege or he would plant Roman camps on his turf. Since Vologeses had a long-standing, iron-clad policy of avoiding actual combat with Roman legions, he stopped challenging the Roman claimant (Tac. Ann. 15.5).
Tacitus claims that Corbulo’s successor as a special commander for the Armenian front, L. Caesennius Paetus, refused to count the installation of a client as military success (Tac. Ann. 15.6.6). According to Tacitus, Paetus “overran what he was not able to hold,” foreshadowing the campaigns of Trajan, and certainly outran his supplies. As Corbulo knew, it was also long-standing, iron-clad Parthian custom to raid wealthy Syria at any sign of Roman weakness. When Corbulo got the news that Paetus had gone back to a winter camp in which he had not invested sufficient initial engineering, he reinforced the bank of the Euphrates, set out a series of lashed boats as artillery platforms to ward off the Parthian horsemen already gathering there, and built a bridge for seizing the offensive. Of course, Vologeses backed off where Romans were prepared and went for Paetus’s camp.
Paetus held to his own long-standing custom of not thinking and sallied. Emboldened by a typical Parthian withdrawal from a minor skirmish, he made a bigger investment the next day and was run off the field. Corbulo was again in no hurry to launch a rescue; he waited for a second note from Paetus urging him to save the standards of his legions. Corbulo would not move into the theater before organizing a camel-train of food. Tacitus claims that Paetus’s men could hardly bring themselves to leave their tents and were thinking of the famous disasters of the Roman Republic at the Caudine Forks (Livy 9.6.1) and Numantia (App. Hisp. 46–7).
Subsequent Roman-Parthian combat was on a broader strategic and even political front but continued in the same vein. Trajan’s massive invasion in 114 was an unprovoked, impossible attempt to take, hold, and reshape the region as part of Rome’s Mediterranean empire. The superb engineering units of Trajan’s army built ships near Nisibis and moved them to the Tigris. They were able to assemble the bridge while under Parthian missile fire (Dio 68.26.2).8 His attempt to hold Parthian territory, as opposed to merely overrunning it, would not have been any more permanent had Trajan not died in winter quarters in 117. Trajan had attempted what might now be called “regime change” in his ambition to replace Parthian dynasts by Roman governors throughout the fertile crescent and the Parthian homelands. It became clear that the Roman Empire could not support the continuous economic and military investment it takes to hold territories against the will of their long-term inhabitants. Septimius Severus’s more modest security goals might have done more damage to the Parthian dynasty generations later.
As noted above, Marcus Aurelius directed his personal attention to the Danube front, leaving the Parthians to Lucius Verus. They had not changed, and in 162 another Vologeses enveloped a Roman legion in Armenia and killed them with missile weapons. Verus sent out forces which were able to pursue the typically retreating Parthians all the way to Ctesiphon, destroying Vologeses’s palace in 165. So those Roman forces also took Ctesiphon, but brought the plague back with them (Dio 71.2).
When Septimius Severus launched an advance across the Euphrates in 198, he did not suffer from the delusion that he could hold much Parthian-dominated territory. Revenge was an honorable motive among Romans, and this invasion might reasonably be described now as retaliation with the aim of deterrence (cf. Germanicus’s “revenge, honor, and terror” directed at the Germans; see above, pp. 238–41). Dio comments on how well-constructed Septimius Severus’s boats for the Euphrates were. He introduced a new practice by not only overrunning Ctesiphon but massacring its inhabitants. In keeping with his strategic aims, he assaulted Hatra too, a site which had also given Trajan trouble (Dio 76.9; for sieges see below). Severus’s province of Mesopotamia was confined to the region of Assyria; he did not try to hold anything east of the Tigris or south of Syria’s frontier on the Euphrates, that hot border in the Cold War, although his punitive invasion extended deeply into the fertile crescent. Some modern scholars have guessed that his sweep of the region, which cost the Parthians prestige, economic assets, and population without a long-term Roman investment enabled the rise of the Sassanid Persians who supplanted the Parthians a generation later. Thus Severus’s focus on revenge per se was the superior strategy. Yet even that “shield” for Syria was too much, Dio predicted. He claims that Severus thought his acquisition a great territory, but it promised to be a cause of war, loss, and entanglement with strange, neighboring peoples (Ep. 75.3).
Romans and Parthians engaged in warfare nearly as ritualized as the Greek or Roman “pitched battle” with infantry. Parthian behavior at the strategic level matched that at the tactical level. At the tactical level, they would respond by returning and picking off, if their retreat were incautiously pursued. At the strategic level, they would retreat, harassing, waiting it out. A generation later, in 217, Macrinus even managed to entrap Parthian cavalry, including camels, in a field of caltrops by using a false retreat (Herod. 14.15.1–3).
SIEGES
Since insurgents rarely opposed the advance of Roman legions openly, the alternative to a Long War against an elusive enemy was often the siege. Roman commanders and armies undoubtedly preferred that alternative. Sieges were also desirable in that they offered opportunities to kill combatants, terrorize populations, eroding their will to resist, and acquire strongholds all in one efficient operation with safely massed forces.
Masada and Jerusalem are rightly famous as sieges, although they are hardly representative operations, given the special difficulties they presented in terrain and extent and sophistication of fortifications. The towns of Galilee and environs offer examples of combat among municipal populations, illustrating the importance even there of small unit cohesion, engineering, integration of cavalry and missile forces, adaptability under extreme conditions, and command and communication in rapidly evolving situations. Actual sieges were often preceded by a campaign of terror, which must have presented municipal populations with a stark if hasty choice as Romans approached. Before Vespasian even set out from Ptolemais with the main column in 67, Placidus had overrun Galilee, killing all he encountered, including the elderly left behind as people fled. Vespasian then halted his column right at the frontier of Galilee to let the population contemplate the possibilities. Placidus found that the more warlike were gathering at Jotapata, so that is where Vespasian went first (Joseph. BJ 3.110–111).9
Just as Vespasian’s advance offered the most detailed model of that phase of war-in-theater, Jotapata supplies the most inclusive study in siege warfare. Vespasian sent forces ahead to surround the city to prevent escapes. He operated typically in constructing a camp on a hill less than a mile away but out of missile or sally range. The sight of a fortified Roman camp was itself terrifying. Having decided to make an example of Jotapata, Vespasian saw no need to proceed straight to an assault. The Roman custom of entering combat thoroughly prepared, rested, and adhering to plan worked well at sieges. Jewish combatants remained in a camp outside the town to protect it; five days’ worth of sallies kept the Romans from simply assaulting the walls.
Vespasian called a planning council. They decided on the classic technique of felling trees and gathering stones to build up to the top of the wall’s shortest section. Construction units built props for a cover against missiles from the walls to protect men composing the ramp. A chain of men kept feeding dirt forward for the project. When the defenders dropped large rocks on the shelter, Vespasian used the big artillery to drive the defenders off the wall. Catapults fired missiles, stone throwers flung large rocks, and burning brands were hurled. Arab bowmen and artillery swept the top of the wall with arrows. The defenders responded with small, stealthy sallies to attack the shelter. Vespasian decided to leave no openings for missiles from above and linked the sections of shelter.
Typically attack spurred innovation, which sparked more innovation in turn. The defenders stretched oxhide screens deflecting missiles while they built the walls higher and added towers. Emboldened, their sallies set some artillery on fire. Then they found out why Romans typically wanted to be on a high point some distance away. Watching for targets, Romans aimed artillery at the site in town at which the evening water ration was issued. Some messengers sent out to get help and supplies made outside contact, but Roman pickets caught on to the routes in the ravines and stopped the rest.
As the mound was completed, Vespasian called up the rams, which were suspended from a framework, not handheld. Artillery moved forward first to drive defenders off the walls. Defenders lowered stuffed bags in front of the rams; Romans attached their standard-issue reaping hooks to poles and cut the ropes to the bags. As the rams were shattering the hastily built extension on the wall, defenders hurled every piece of wood they could light on fire and set both the attacking pieces and the protective screens on fire. Legions farther back covered their equipment and even buried it. Vespasian himself was scraped on the foot by an arrow and then had to show himself prominently to prevent panic among Romans. As night fell, the defenders were outlined by the fires behind them, so they were getting the worst of the exchange of missiles, a common problem in sieges.
Dawn came up as chunks of wall fell out, and defenders tried to fill the gap with their bodies and equipment. Uncharacteristically, Vespasian allowed his forces only a short rest from the night before. He dismounted “picked cavalry” with their armor and long javelins to go up gangways to the gaps first with specially selected infantry behind them. Cavalry circled the operation to catch any fugitives who came pouring out. Slingers, bowmen, and artillery were standing by in support. Other men were to go over siege ladders, assuming that defenders would come off the walls to fight in the town once Romans had entered a breach and were behind them. All the trumpeters sounded the attack; clouds of arrows flew into Jotapata. Josephus is undoubtedly not overstating the terror among the population. No defenders believed that they could survive; they charged out as Romans tried to enter. At least one Roman unit was broken by flaming oil which ran under Roman armor, even in testudo formation, but fresh Roman units kept coming forward.
Clearly the narrowness of the gaps was a problem, as was to prove true at Jerusalem and for Septimius Severus at Hatra. Engineering units came into play even in mid-engagement. Vespasian ordered construction of three iron-clad towers manned by missile forces, knowing that they would be hard to set on fire or to turn over. Josephus claims that Jotapata held out for an additional forty-seven days. Finally, as so often, a traitor told Vespasian that the defense was exhausted; the sentries slept. Titus, Vespasian’s son, led Roman units over the walls on ramps which had been built up all that time in the depths of the nights. Romans killed everyone except for women and babies, spending days searching the sewers and caves. Some of the defenders committed suicide; some held out to the end in one tower. Josephus estimated forty thousand Jewish dead (Joseph. BJ 3.141–340). This operation offers a compendium of siege and counter-siege tactics, and Josephus’s account is a daunting picture of the maiming and death on each side (figure 11.2).
Siege warfare in this theater terrorized populations and led many towns to reject or betray resistance, reducing the number of havens. It also simply reduced the numbers of the resistance, catching them where they were gathering to organize. As the will to resist was broken, Romans could acknowledge situations in which resistors occupied a town against the will of its residents. At Tarichaeae, whose unfortified side was on a lake, Titus, his engineers, and legions displayed versatility by building rafts to pursue fugitives in a unique small craft battle. Sometimes, as at Tarichaeae, Titus and Vespasian tried to distinguish residents from those who had entered recently (Joseph. BJ 3.462–542). Most towns and fortresses subsequently surrendered; many had holes knocked in their walls anyway, since increasingly desperate resistance might have occupied the sites after Roman forces had moved on. This demonstrates that sieges were not simply tactical operations at one site but nodes in grand tactical campaigns to “pacify” theaters, and each siege in theater would change the next for both assailants and besieged.
Figure 11.2 Celebration of the triumph of Titus and his victory in the Great Jewish War, a panel from the Arch of Titus, Rome, ca. A.D. 82. Photo Credit: L. Tritle.
The pattern of Jotapata is visible again at Gamala, even where the town itself was nearly vertical on a camel-backed ridge, although this operation also offers rarer snapshots of Roman vulnerabilities in urban combat. Vespasian was unable to encircle the city first because of its site, but he could put the standard camp on a nearby hill which overlooked it. He filled in ravines and trenches in order to start constructing ramps at the complex site. He moved engines up; artillery drove defenders off the wall; rams opened the wall; trumpets sounded. As at Japha (Joseph. BJ 3.302–303), Romans met dangerous resistance on narrow streets, while objects could be thrown down from above. Urban warfare was even worse at Gamala, where Romans had to force their way up extraordinarily steep streets against the sheer weight of defenders. They tried to move on the roofs of houses which collapsed under heavy Roman armor. Collapsing houses offered a supply of stones for defenders. It was nearly a failed siege; Romans had to leave and regroup. Many residents fled, knowing that the very success of the defense deterred any offer of clemency. The operation finally succeeded via Roman undermining, which often simply removed stones from the base of walls instead of dangerous tunneling (a mere three soldiers rolled five stones out of the base of one tower [Joseph. BJ 4.4–83]).
Titus’s subsequent operations at the extremely complex site of Jerusalem were troubled by familiar factors: the influx of criminals, desperate refugees, ideological resistance, local factionalism whose effects could be for or against Roman interests. In spite of the scope and scale of the defenses, Vespasian had encircled the site with camps, some at cities like Gerasa which had to be taken to complete the circle. He meant to ensure that no one left without surrendering himself. Titus adapted to the great scale by moving camps, constricting the circle as the nature of operations changed, and using information from the captured to direct artillery to sites within. He also repeated errors in sending men through too small a breach in the second wall into the New City. Only a covering fire of arrows got him and his men out safely to regroup (Joseph. BJ 5.333–341). Rapidly adapting defenders even undermined Roman platforms by tunneling (Joseph. BJ 5.468–471, 302–521). Roman legions could suffer greatly even in this sphere of warfare in which they most excelled.
Hatra, the caravan city on the Arab-Persian frontier, became notorious for testing the limits of Roman siege warfare. Dio is oddly dismissive of Hatra’s size and importance, given the wealth collected in offerings to its great Sun temple (cf. 68.31, 76.12.2). There was very little potable water locally and no wood, ruling out building operations like Vespasian’s. There were also insects which probably bore disease (Dio 68.31; Hdn. 3.1.2–3 may be confused with the hurling of jars of stinging insects). Dio claims that Trajan simply assaulted the walls with cavalry. The locals were particularly good archers who also had some sort of ballista or scorpio which could deliver a very high rate of fire, presumably making it nearly impossible to approach the walls, as well as some naphtha-based flame-throwing device, which must be what destroyed Severus’s artillery later.
Severus did breach a wall at one point, and supposedly his troops became outright mutinous when he would not exploit it, but the experiences of Vespasian and Titus justify his position. His troops’ anger presumably resulted from the very high casualties among Roman forces combined with the lost opportunity to loot a shrine city. The chronology is much debated, but if Severus attacked Hatra on his way east, failing and sweeping by it to the south, that illustrates that sieges could have objectives other than holding territory or terrorizing populations. If he attacked it for either the first or the second time during his withdrawal from Mesopotamia, that would indicate that he was aiming at the Sun’s treasury and, if it were the second time, to recover from a blow to his military prestige. Jotapata, Hatra, and Gamala all lead to the topic of the next section.
DEFEAT AND RETREAT
Morale issues could be cumulative, witness mutinous riots against Germanicus in 14, a mass phenomenon perhaps combined with a conspiracy to desert or to kill officers. The complaints were those which would surface repeatedly during the Empire, especially on the northern frontiers: excessive length of service before demobilization; insufficient rewards. To demonstrate the personal costs of such long service, men pushed Germanicus’s hands into their mouths to feel their missing teeth and ripped off clothing to show the scars from wounds and flogging (which cannot, therefore, have been too shameful). Germanicus wisely rejected mobilizing auxiliary forces against legions, but asking the legions to turn on hard-core mutineers in their own small units had horrible results. Tacitus describes their subsequent fervor to set out against Germans as atonement for their earlier madness (Ann. 1.29–49).
It was these legions, already mobilized for a campaign of terror against German villages, which Germanicus led on a detour in the next year to the site of the massacre of Varus in the Teutoburg forest where Romans had been left unburied. His units included men who had escaped the slaughter or captivity among the Germans. In a pointed contrast to Varus, Germanicus entered the marshes preceded by scouts, bridge builders, and builders of camps. They found not only Varus’s camp but Roman bones attached to trees and what they identified as altars at which Roman officers had been sacrificed.
Tacitus reports criticism of Germanicus’s participation in the burial project, both because Roman forces might have been demoralized and terrified by the sight of the Roman dead and because Germanicus should not have been handling any funerary articles as an augur (Ann. 61–2). Roman literature more conventionally features the laments and dire fates of Roman opponents, but Tacitus intended his depiction of the aftermath of civil warfare in Italy to be affecting when he offered another scene of contemplation of Roman dead. At the site of the first battle of Cremona, bodies of the defeated, some, Tacitus claimed, dismembered, lay in heaps together with horses. Tacitus depicts Vitellian officers as offering a battlefield tour of what each unit did. Tacitus clearly approved of the “few” observers moved to tears by the sight of mounds of Roman dead, but Vitellius supposedly did not have the sensibility to avert his eyes (Hist. 70–71).
Casualties, of course, could occur even in winning engagements, as the siege at Jotapata demonstrated. Trajan’s Column, commemorating victory in wars in Dacia, admits to Romans wounded in combat. It exhibits a slumped soldier, propped up by a comrade, reaching out to a medic and a legionary with a steely expression having a leg bandaged by another medic (Lepper and Frere 1988: casts 102–3). Dio claimed that Romans ran out of bandages at Tapae, so that Trajan had to rip up his own clothing. He then honored the losses by establishing an altar and annual sacrifices on the spot (68.8.2). Some casualties probably occurred in the aftermath of battle, when fates contrasted most starkly. The column juxtaposes Trajan’s distribution of awards to auxiliaries with bound Roman captives, tortured by Dacian women (Lepper and Frere 1988: casts 115–16, 117 respectively). This may be connected with the Roman posting of Dacian heads on poles outside a stone fort later and heads offered Trajan as trophies (Lepper and Frere 1988: casts 140, 183). Pride in this head-hunting suggests some degree of desensitization to dismembered bodies as well as supporting a recent argument that Roman military sensibilities in particular did not call for the self-control or restraint in self-assertion lauded by philosophically inclined texts. Insult was likely to lead to vengeance (Lendon 1997: 247).
Others reading this literature have looked for some evidence for Roman post-traumatic stress disorder, given the horrible intimacy of death inflicted in legionary combat, not to mention the perpetration of campaigns of terror against civilian populations. These included rage-enabled explosions of violence and also directed extermination of populations carried out systematically over days, extended scenes of which are distressing to modern military personnel even in print. Highly personal violence and ambiguity as to who is or is not a combatant are factors which increase the prevalence of PTSD (Garbutt 2006: 21). Yet the main indication of PTSD may be the very numbness in the aftermath of battle for which Tacitus expresses disdain. PTSD routinely involves anger, lack of concentration, “overall numbing of responsiveness” as well as “a marked disinterest in important activities, feelings of either detachment or alienation.”10 Confusingly, numbness could alternatively have been a symptom of pathological depression, if “only” situational depression in some cases, for example, perhaps, Crassus in the Republic, unable to come out of his tent in the aftermath of Carrhae. Instead of reassuring his men, still trapped deep in hostile territory after Carrhae, after the death of his son, he lay in his tent alone with his face covered. Subordinates were unable to rouse him and held a council of war and issued marching orders on their own (Plut. Cras. 27; depression is frequent post-combat, yet depression and PTSD reflect two different anomalies of brain chemistry).
Both reactions violated cultural norms among the Roman elite who eschewed the restraint recommended by Greek philosophers. Instead they valued a sensibility open to the grief of others, so Tacitus’s disdain for those inured to violence would have been widely shared (MacMullen 1980: 254–5). Germanicus’s battlefield tour as well as that of Vitellius might have counteracted psychological trauma rather than worsening it. Debriefing (analyzing the sequence of events in detail for others) sometimes allows individuals to process events through group discussions when their own mental functions are too traumatized. Such debriefings operate as rituals moderating emotion and lending meaning to events; they make the military unit a source of meaning again (i.e., as long as no blaming occurs; cf. Shalev 1991: 5).
Later in that eastern theater, Paetus’s men could not be lured into another ambush by the Parthians. They would hardly leave their tents except to take their turns on the walls. Tacitus claims that some were simply very obedient to orders, while others were cowards. He adds that Paetus himself was “unable to overcome adversity” and began to neglect all military duties, perhaps with a pathological degree of detachment (Tac. Ann. 15.15). It is, nonetheless, oddly impressive that Paetus built a bridge for the route for his own retreat. Parthians walked into the Roman camp before they had even evacuated it and picked out Roman-acquired loot. Tacitus concedes that some stories might have been invented later to heap disgrace on Paetus, yet repeats the claim that Romans were also subjected to marching under a yoke. He vouches for the fact that Paetus marched back at forty Roman miles per day, abandoning wounded as he went, which Tacitus considered just as bad as fleeing a battlefield. Corbulo had pointedly met them without any of the display of arms which Roman legions might expect to greet them on return from combat, but his men, Tacitus claims, saw their fellow soldiers through tears, a suitably sensitive reaction to their plight (Tac. Ann. 15.16–17).
Paetus’s soldiers evoke the topic of infamia, disgrace, and the need for some redemption via piaculum, an expiation. Tacitus depicts Roman commanders as studying the minds of their men (e.g., militum animi in Ann. 2.12.3). Experience of or even the fear of disgrace certainly influenced behavior deeply. An individual considered a coward might commit suicide (Tac. Hist. 2.30). Indiscipline, even that inspired by powerful grievances, might later be experienced as an unbearable disgrace, as in the case of Germanicus’s mutineers (Tac. Ann. 1.3, 49). The greatest disgrace was defection to an enemy, a lesser people, and legions drawn into revolt by the Batavians and Gauls on the Rhine suffered not only infamia but conscientia flagitii and ignominia. Roman commanders protected them from insult by other troops, perhaps fearing something like a mass depression or even an attempt at group expiation as non-rebels killed them to recover unit reputation (Tac. Hist. 4.62, 72).
Motivation for enduring combat has often been studied since World War II, and the standard claim of the literature into the twenty-first century is that men endure combat because they will not abandon the other men in their units. That means that unit cohesion is not simply an outcome of combat or of training but the primary motivator in a tactical encounter. Even when ideological motivations are present, unit cohesion is the sole motivator which does not erode under the stress of combat (Steckel 1990: 300). In fact, cohesion is, in general, enhanced by group experience, especially by group success, that is, victory helps. In spite of recent studies which attempt to divide unit cohesion into task cohesion and social cohesion, it now appears that task cohesion builds social cohesion.11 Unit cohesion is one of the best protections against PTSD since it enhances psychological resilience in general (Garbutt 2006: 29). This cumulatively suggests that Romans largely achieved resilience and avoided collapse at the strategic level after a tactical or even grand tactical defeat through a synergistic interaction of factors including a military culture of adaptability to local tasks, unit cohesion, and prior success.
BIBLIOGRAPHY
Arrian. 1885. Scripta minora. Ed. R. Hercher. Leipzig.
Ben-Shalom, U., Z. Lehrer, and E. Ben-Ari. 2005. “Cohesion during military operations: A field study on combat units in the Al-Aqsa intifada.” Armed Forces & Society 32: 63–79.
Bosworth, A. B. 1977. “Arrian and the Alani.” HSCP 81: 217–55.
Campbell, B. 1984. The emperor and the Roman army 31 B.C.–A.D. 235. Oxford.
———. 1987. “Teach yourself how to be a general.” JRS 77: 13–29.
———. 1993. “War and diplomacy: Rome and Parthia, 31 B.C.–A.D. 235,” in Rich and Shipley, 213–40.
———. 2002. War and society in imperial Rome 31B.C.–A.D. 284. London.
Dixon, K. R. 1992. The Roman cavalry from the first to the third centuriesA.D. London.
Eadie, J. W. 1967. “The development of Roman mailed cavalry.” JRS 57: 161–73.
Garbutt, P. F. C. 2006. The paradox of fight or flight—A leadership guide to understanding and mitigating operational stress injuries. Ottawa.
Goldsworthy, A. K. 1996. The Roman army at war 100 B.C.–A.D. 200. Oxford.
———. 2003. The complete Roman army. London.
Hassall, M. W. C. 1970. “Batavians and the Roman conquest of Britain.” Britannia 1: 131–6.
Isaac, B. 1992. The limits of empire: The Roman army in the east. Rev. ed. Oxford.
Lee, A. D. 1996. “Morale and the Roman experience of battle,” in A. B. Lloyd (ed.), Battle in antiquity. London, 199–217.
Lendon, J. E. 1997. Empire of honour. The art of government in the Roman world. Oxford.
———. 2005. Soldiers and ghosts: A history of battle in classical antiquity. New Haven.
Lepper, F., and S. Frere. 1988. Trajan’s column. A new edition of the Cichorius plates, Gloucester.
MacMullen, R. 1980. “Romans in tears.” CP 75: 254–5.
———. 1984. “The legion as a society.” Historia 33: 440–56.
Mattern, S. 1999. Rome and the enemy: Imperial strategy in the principate. Berkeley.
Pitts, L. F. 1989. “Relations between Rome and the German ‘kings’ on the middle Danube in the first to fourth centuries A.D.” JRS 79: 45–58.
Sabin, P. 2000. “The face of Roman battle.” JRS 90: 1–17.
Shalev, A. 1991. Historical group debriefing following combat. Frederick.
Shaw, B. D. 1984. “Bandits in the Roman empire.” P&P 105: 3–52.
Steckel, F. 1990. Morale and men. Ann Arbor.
Thorne, J. T. 2007. “Battle, tactics, and the emergence of the limites in the west,” in P. Erdkamp (ed.), A companion to the Roman army. Oxford, 218–34.
Wong, L., T. A. Kolditz, R. A. Millen, and T. M. Potter. 2003. Why they fight: Combat motivation in the Iraq war. Carlisle.